The recent rise in the availability of sensors to monitor our day-to-day lifestyles calls for techniques to glean information out of the generated data. For many of these sensors, the data recorded is periodic in nature. Moreover, it has been shown that a change in this periodic data is correlated to a shift in physical or mental health. We present a technique to compare this repetitive data, both qualitatively and quantitatively. Our method first summarizes the sensor measurements collected from smart homes of elderly into textual statements. Dissimilarity between the text summaries is then computed to quantify the comparison.